Automatic Modulation Classification (AMC) is an approach to identify an observed signal's most likely modulation scheme without any a priori knowledge of the intercepted signal. In this research, the authors present a new direction for both stages of feature-based (FB) approach. In the feature extraction stage, the authors design a new architecture that 1) removes the bias issue for the estimator of fourth-order cumulants, and 2) extracts polar-transformed information of the received IQ symbols, and finally 3) forms a unique dataset to be used in the labeling stage. Furthermore, the authors contribute to increasing the classification accuracy in low signal-to-noise ratio (SNR) conditions by employing the deep belief network (DBN) platform in addition to the spiking neural network (SNN) platform to overcome execution latency concerns associated with deep learning architectures. For this research, the authors first study each individual FB AMC classifier to derive their respective upper and lower performance bounds and then propose an adaptive framework that is built and developed with these findings. This framework aims to efficiently classify the modulation scheme by intelligently switching between these different FB classifiers to achieve an optimal balance between accuracy and execution latency for any observed channel conditions derived from the main receiver's equalizer. Subsequently, a performance analysis is conducted using the standard RadioML dataset to achieve a realistic evaluation. Numerical results indicate a notably higher classification accuracy, by 16.02% on average, when DBN is employed, whereas SNN requires significantly lower execution latency to label the modulation scheme when compared against two other modified FB classifiers that are built upon convolutional and recurrent neural networks, shown to be reduced by 34.31%.

The procedure of classifying a detected signal’s modulation scheme with no a priori information is known as automatic modulation classification (AMC). AMC has presented numerous contributions toward civilian and military applications. However, in the scientific literature, designing automatic modulation classifiers has been limited to a single particular simulated environment in most cases of study. Hence, the performance of such classifiers lacks: 1) reliability in real-world environments and 2) multivariate environment analysis for AMC operation in dissimilar surroundings. These two reasons represent a significant obstacle to the real-world implementation of such classifiers. Therefore, in this research, we present our contribution to remove these obstacles by generating an emulated signal reference dataset named MIMOSigRef-SD. It includes a wide variety of signal streams modulated by different digital modulation schemes, such as M-QAM, MIL-STD-188-110 B/C standard-specific QAM, M-PSK, M-APSK, DVB-S2/S2X/SH standard-specific APSK, and M-PAM with different modulation orders, each with different multiple-input multiple-output (MIMO) system configurations in order to provide an extensive signal reference. Each signal is also exposed to the realistic impact of different channel models at 2450 MHz, specifically Vehicular-A/B, Pedestrian-A/B, and the Butler, through a channel emulator. This signal reference includes the randomly generated transmit signals, characteristics of the emulated environments, and the recorded received symbols. This enables the dataset to be highly applicable to applications beyond AMC as well. In general, this signal reference can help analyze, evaluate, and design any other radio frequency transceiver tasks under realistic environmental effects. We also validate the elements of the signal reference by comparing the emulation and simulation bit error rate of the designed communication system.

Railyards are one of the most challenging and complex workplace environments in any industry. Railyard workers are constantly surrounded by dangerous moving objects, in a noisy environment where distractions can easily result in accidents or casualties. Throughout the years, yards have been contributing 20–30% of the total accidents that happen in railroads. Monitoring the railyard workspace to keep personnel safe from falls, slips, being struck by large object, etc. and preventing fatal accidents can be particularly challenging due to the sheer number of factors involved, such as the need to protect a large geographical space, the inherent dynamicity of the situation workers find themselves in, the presence of heavy rolling stock, blind spots, uneven surfaces and a plethora of trip hazards, just to name a few. Since workers spend the majority of time outdoors, weather conditions also play an important role, i.e., snow, fog, rain, etc. Conventional sensor deployments in yards thus fail to consistently monitor this workspace. In this paper, the authors have identified these challenges and addressed them with a novel detection method using a multi-sensor approach. They have also proposed novel algorithms to detect, classify and remotely monitor Employees-on-Duty (EoDs) without hindering real-time decision-making of the EoD. In the proposed solution, the authors have used a fast spherical-to-rectilinear transform algorithm on fish-eye images to monitor a wide area and to address blindspots in visual monitoring, and employed Software-Defined RADAR (SDRADAR) to address the low-visibility problem. The sensors manage to monitor the workspace for 100 m with blind detection and classification. These algorithms have successfully maintained real-time processing delay of ≤0.1 s between consecutive frames for both SDRADAR and visual processing.

Human-in-the-Loop (HuIL) systems are systems that consider presence of human, human errors and human interactions with a system. Accurate and validated sensing, coordinated control and decision over human errors and interactions, even at a scale and complexity should be done reliably and seamlessly. In these systems, sensors, while being discretized and processed digitally, nevertheless represent physical properties. Therefore, modeling of these very large systems faces two key challenges: a) while computing aspects can be represented using discrete-time events, physical parameters can only be monitored adequately using continuous-time models, and b) with human factors involved, behavior modeling is intractable and highly inaccurate. Conventional and state-of-the-art systems always limit themselves in terms of the bounded measure calculated from the data available, without factoring in human rational information verification. In this paper, uncertainty theory-based cyber-physical state-space approximation is proposed. These approximations are further morphed topographically based on Ricci Flow and Fuzzy qualifiers to match human verification of sensor data. The results show that sensor information can be formally approximated, verified and identified by human belief degrees. {To the best of our knowledge this is the first implementation of Ricci flow in the domain of Fuzzy CPS, and it has the potential to improve overall reachable set interpretability with qualitative and quantitative interpretation.